SAMPLITUDE RULES (DAW Summing)

[stefancrs]

1st. What they said about brownian molecule motion is that the accuracy 32 bit floats provide are greater than the effect of brownian air molecules motion. However this is for ONE sample, if truncation or loss of accuracy happens in every calculation even with a "low" sampling frequency of 44.1 khz that 44.100 truncations a second multiplied by the numerous places your signals flows to and fro and ......
Can you say "Exponential Growth" ? The differences could probably manifest as audible artefacts.

2nd i did not say that the summing procedure is the only reason but also both objective stuff : e.g bad PDC timing giving phasing problems, panning laws and subjective stuff : e.g. : logic 5.5.1 that is use has 127 step values in it's mixer - the volume i need may not be in those numbers.

There's no such thing as bad PDC. Either it's there or it isn't. If it's buggy (which I doubt any hosts pdc is) it most likely gets fixed quite rapidly. Which hosts does not allow the user to set the panning law? Does logic not use floats for the volumes in the mixer (I don't know, I don't use logic, but if it only uses 127 steps I wouldn't want to use it either )

What if one host miscalculates the latency of a plugin. From what i remember not all of the plugins report their latency and some even report it incorrectly.

Hosts does not calculate the latencies of plugins. Ever. If a plugin reports a faulty latency, it will affect all hosts with PDC. If they don't report latency, that too will affect all fully PDC enabled hosts the same way. The same way.

As for the logic bit, why do you think i want another app to mix :p. I asked if it is only 127 steps cause from what i could gather looking everywhere it was, and the reply i got was that this was indeed correct. Bummer as wouldn't change logic with anything for my midi stuf

Why not use your own volume control instead then? Are you sure you're not refering to automation via midi? You'd easily hear each step in some cases if the volume control is only 7bit.

[popsych]

1st. What they said about brownian molecule motion is that the accuracy 32 bit floats provide are greater than the effect of brownian air molecules motion. However this is for ONE sample, if truncation or loss of accuracy happens in every calculation even with a "low" sampling frequency of 44.1 khz that 44.100 truncations a second multiplied by the numerous places your signals flows to and fro and ......
Can you say "Exponential Growth" ? The differences could probably manifest as audible artefacts.

Spectacularly ignorant rubbish. Give me a link or a refence to any piece of research even claiming to show that amplitude differences are cumulative over time.

How do you propose to prove your claim? Put up or shut up.

Ignorant rubbish ? Do you even know how many times each sample value has something done to it in every cycle ? Plugins are not just one addition/division/multiplication/subtraction. I'll tell you what - The excellent Voxengo, Kjaerhus, Izotope (in that order :p) all work in 64 bit. Why do you think that is.

I don't have to prove my claim to you : I have the better mixes (IMHO) not you - why should i give you the benefit.

I could have said shut up right back at ya but u seem such a d*** i can't be bothered to

[nuffink]

1st. What they said about brownian molecule motion is that the accuracy 32 bit floats provide are greater than the effect of brownian air molecules motion. However this is for ONE sample, if truncation or loss of accuracy happens in every calculation even with a "low" sampling frequency of 44.1 khz that 44.100 truncations a second multiplied by the numerous places your signals flows to and fro and ......
Can you say "Exponential Growth" ? The differences could probably manifest as audible artefacts.

Spectacularly ignorant rubbish. Give me a link or a refence to any piece of research even claiming to show that amplitude differences are cumulative over time.

How do you propose to prove your claim? Put up or shut up.

Ignorant rubbish ? Do you even know how many times each sample value has something done to it in every cycle ? Plugins are not just one addition/division/multiplication/subtraction. I'll tell you what - The excellent Voxengo, Kjaerhus, Izotope (in that order :p) all work in 64 bit. Why do you think that is.

I don't have to prove my claim to you : I have the better mixes (IMHO) not you - why should i give you the benefit.

I could have said shut up right back at ya but u seem such a d*** i can't be bothered to

Another fantasist. Talks the talk 'til challenged and then flounces off with a "Nuh nuh. I'm better than you."

[bmanic]

Whoah! This thread turned into total voodoo. Guys, there are MUCH bigger factors affecting sound than summing algos. Come on, do some ABX testing on yourself and prepare to be humbled!

stefancrs, om du nångång kommer på besök till finland så PM mig och kom på ett stop kall god finsk öl!

Cheers!
bManic "finnen som talar svenska"

[stefancrs]

Whoah! This thread turned into total voodoo. Guys, there are MUCH bigger factors affecting sound than summing algos. Come on, do some ABX testing on yourself and prepare to be humbled!

stefancrs, om du nångång kommer på besök till finland så PM mig och kom på ett stop kall god finsk öl!

Cheers!
bManic "finnen som talar svenska"

Klart jag ska!

[Omniphonix]

I don't claim to know all the technical bits about software mixers, but I did notice things were sounding better when I moved from Cubase SX 1 to SX 2. I think that was the revision when they implemented 64-bit floating point from input to master out downsampling algo. There's got to be something to it.

[stefancrs]

I don't claim to know all the technical bits about software mixers, but I did notice things were sounding better when I moved from Cubase SX 1 to SX 2. I think that was the revision when they implemented 64-bit floating point from input to master out downsampling algo. There's got to be something to it.

Maybe, but I doubt it, any vst effect would break the 64bit and convert to 32bit. And afaik SX2 only uses 32bit signals. Usually percieved differences in sound is due to small changes in volume.

[bmanic]

Most quality plugins are 64bit internal though. Would that break it to 32bits? Do they dither at the output before the next plugin? I know nothing about software code so that's why I'm asking.

Cheers!
bManic

[stefancrs]

Most quality plugins are 64bit internal though. Would that break it to 32bits? Do they dither at the output before the next plugin? I know nothing about software code so that's why I'm asking.

Cheers!
bManic

All vst plugins must use 32bit float as input and output.

[popsych]

1st. What they said about brownian molecule motion is that the accuracy 32 bit floats provide are greater than the effect of brownian air molecules motion. However this is for ONE sample, if truncation or loss of accuracy happens in every calculation even with a "low" sampling frequency of 44.1 khz that 44.100 truncations a second multiplied by the numerous places your signals flows to and fro and ......
Can you say "Exponential Growth" ? The differences could probably manifest as audible artefacts.

Spectacularly ignorant rubbish. Give me a link or a refence to any piece of research even claiming to show that amplitude differences are cumulative over time.

How do you propose to prove your claim? Put up or shut up.

Ignorant rubbish ? Do you even know how many times each sample value has something done to it in every cycle ? Plugins are not just one addition/division/multiplication/subtraction. I'll tell you what - The excellent Voxengo, Kjaerhus, Izotope (in that order :p) all work in 64 bit. Why do you think that is.

I don't have to prove my claim to you : I have the better mixes (IMHO) not you - why should i give you the benefit.

I could have said shut up right back at ya but u seem such a d*** i can't be bothered to

Another fantasist. Talks the talk 'til challenged and then flounces off with a "Nuh nuh. I'm better than you."

I have no problem doing the test with stefancrs and i will - just to put my money where my mouth is. But you are , as i said not, worth bothering with. If you care that i don't care about you then ...... too bad. I still don't give a tiny rats a**.

OMG LOL YOU GUYS ARE EVEN DUMMER THAN I THOUGHT- DONT YOU KNOW THAT ONLY PROTOOLS IS ADVANCED AND WORKS AT 256 BIT INTERNALLY- YOU CANT EVEN GET CLOSE WITH CUBASE LOL ROFL. SAMPLITUDE ONLYWISHES IT COULD BE HALF AS GOOD AS PROTOLLS VERSION 1 LOL. MY MACITOSH IS 64 BIT AND YOU SORRY BILL GATES WINDOWS SAMPLITUDE CRAP EATS ITS DUST. JUST WAIT TIL MY ALBUM COMES OUT NEXT MONTH LOL ROFL, YOU WILL HEAR REAL RAP PRODUCED ON REAL HARDWARE LOL ROFL.

Yeah but Pro tools 1 had the 256 bit only in it's mixer. I hear the new sonar will upsample to 500000000000000 Mhz 1024 kbits do mixing and then it will beat the ass off your PT system i tell ya

Most quality plugins are 64bit internal though. Would that break it to 32bits? Do they dither at the output before the next plugin? I know nothing about software code so that's why I'm asking.

Cheers!
bManic

In a word - Yes they dither down.

[nuffink]

I have no problem doing the test with stefancrs and i will - just to put my money where my mouth is. But you are , as i said not, worth bothering with. If you care that i don't care about you then ...... too bad. I still don't give a tiny rats a**.

You're a fantasist mate. We see them all the time here. You won't do any test. You'll wriggle out of it or just disappear. If by some slim chance a test does come out of this you'll fail it, whinge that it wasn't fair and then disappear.

Same as it ever was.

[popsych]

I have no problem doing the test with stefancrs and i will - just to put my money where my mouth is. But you are , as i said not, worth bothering with. If you care that i don't care about you then ...... too bad. I still don't give a tiny rats a**.

You're a fantasist mate. We see them all the time here. You won't do any test. You'll wriggle out of it or just disappear. If by some slim chance a test does come out of this you'll fail it, whinge that it wasn't fair and then disappear.

Same as it ever was.

Now if you weren't such a d*** and if i had feelings i would have been offended.

Stefan- i've pmd you so we can setup the test. Reply when you can so we can arrange it.

[Omniphonix]

I don't claim to know all the technical bits about software mixers, but I did notice things were sounding better when I moved from Cubase SX 1 to SX 2. I think that was the revision when they implemented 64-bit floating point from input to master out downsampling algo. There's got to be something to it.

Maybe, but I doubt it, any vst effect would break the 64bit and convert to 32bit. And afaik SX2 only uses 32bit signals. Usually percieved differences in sound is due to small changes in volume.

You're right, it is 32-bit.

Improvements such as the new VST audio engine with 32-bit floating point audio resolution,

[dasdeck]

fengshui ?? me?? don't get it regarding to my post .

Any def musicians posting here? Maybe they can confirm that samplitude, SX, Fruity and other soft do NOT sound different.

DECK

[popsych]

fengshui ?? me?? don't get it regarding to my post .

Any def musicians posting here? Maybe they can confirm that samplitude, SX, Fruity and other soft do NOT sound different.

DECK

I think he was implying that you are "hearing" stuff.